Peroxisomes are ubiquitous components of eukaryotic cells. Metabolic functions of these organelles are widespread and include hydrogen peroxide based respiration, β oxidation of very long chain fatty acids (VLCFA), and biosynthesis of ether phospholipids, bile acids, and isoprene compounds.¹ Several human diseases are caused by mutations in genes encoding peroxisomal metabolic enzymes,² but the most dramatic loss of peroxisome function is observed in the peroxisome biogenesis disorders (PBD; MIM 601539).³

For PBD two broad clinical spectra can be distinguished, the Zellweger spectrum, accounting for about 80% of all PBD patients, and the rhizomelia chondrodysplasia punctata (RCDP) spectrum.^3,4 The Zellweger spectrum consists of three overlapping clinical phenotypes that represent a continuum of disease severity, including Zellweger syndrome (MIM 214100) as the prototype and most severe example of this group, neonatal adrenoleucodystrophy (NALD, MIM 202370) as an intermediate form, and infantile Refsum disease (IRD, MIM 266510) as the mildest variant. Zellweger syndrome patients have characteristic dysmorphic features, severe neurological dysfunction including hypotonia, seizures and poor feeding; they have eye abnormalities like cataracts, liver dysfunction and skeletal defects. They rarely survive the first year of life. Patients with NALD, IRD and atypical Zellweger syndrome have similar but less severe clinical signs and can survive up to several decades. RCDP is clinically and genetically distinctive from the Zellweger syndrome spectrum and includes classical RCDP as the prototype and also milder variants. Patients with classical RCDP have unique clinical symptoms including proximal shortening of the limbs (rhizomelia), cataracts, and profound psychomotor retardation.

PBDs can be caused by defects in any of several processes in organelle formation, including the synthesis of peroxisome membranes, the recognition of newly synthesised peroxisomal matrix proteins, or any of the downstream steps in peroxisomal protein import. Progress over the last two decades has led to the identification of 13 different human PEX genes involved in peroxisome biogenesis, explaining the primary genetic defect of all 13 known complementation groups (CG) for PBD patients. The defective genes are PEX1 for CG1,^5,6PEX5 for CG2,⁷PEX12 for CG3,^8,9PEX6 for CG4,¹⁰PEX10 for CG7,^11,12PEX26 for CG8,¹³PEX16 for CG9,¹⁴PEX2 for CG10,¹⁵PEX7 for CG11,^16,17PEX3 for CG12,¹⁸PEX13 for CG13,^19,20PEX19 for CG14,²¹ and PEX14 for CGK.²²PEX1 mutations are the most common cause and account for two thirds of all PBD patients.

PEX1 maps to 7q21–q22 and encodes peroxin 1 (PEX1), a 147 kDa member of the AAA protein family of ATPases (ATPases associated with diverse cellular activities). PEX1 is classified as a type II AAA ATPase as it comprises two ATPase domains (AAA cassettes) containing two functional domains D1 and D2, each spanning over about 200 amino acids (D1: 198 amino acids (p.436–633); D2: 185 amino acids (p.717–901)) with the second one being highly conserved.²³ Both AAA cassettes contain a Walker A and B motif (Walker A1 and B1; Walker A2 and B2).²⁴ Given that PEX1 has not been crystallised completely, knowledge of the structure is limited.²³ So far, only the structure of the N terminal domain is known. It shows striking similarity to that of two other type II AAA ATPases, namely valosin containing protein (VCP) and N-ethylmaleimide sensitive factor (NSF). VCP and NSF also seem to be involved in maintaining organelle function.

We analysed the PEX1 gene in a consecutive series of 33 patients with Zellweger spectrum. We report the full spectrum of molecular defects in these patients as well as of those who have already been reported. We also describe the clinical features related to PEX1 gene mutations and establish a genotype–phenotype correlation.

METHODS

The study included 168 Zellweger spectrum patients with PEX1 mutations. Thirty three patients were analysed in our laboratory. The study was approved by the local university ethics committee. Genomic DNA from cultured skin fibroblasts or leucocytes was used as template. Genomic DNA was obtained from leucocytes in 50 white controls. Mutations were detected by different methods including single strand conformation polymorphism (SSCP) analyses as a screening technique on the basis of genomic or cDNA, followed by direct sequencing of polymerase chain reaction (PCR) fragments with an abnormal electrophoresis pattern.^5,25,26 It is well known that SSCP analyses fail to detect 10–20% of mutations depending on fragment size and electrophoresis conditions.²⁷ Thus we and others have also used reverse transcriptase PCR (RT-PCR) to amplify the coding sequence of PEX1, followed by direct sequencing,^6,24,28,29 or direct sequencing of PEX1 exons that were amplified by PCR with intronic primers.^30–33 For so far undescribed mutant alleles, polymorphism was excluded by analysing 100 white control alleles.

The mutation nomenclature used follows the recommendations of the Human Genome Variation Society (HGVS, http://www.genomic.unimelb.edu.au/mdi/rec.html).³⁴ All mutations were adjusted to mRNA reference sequence NM_000466 (version: NM_000466.1; source sequence: AF026086; version: AF026086.1; protein product: NP_000457; version: NP_000457.1).

RESULTS AND DISCUSSION

Mutation data analyses from 168 patients with 288 described mutant PEX1 genes revealed 46% missense mutations, 46% insertions and deletions, 3% splice site mutations, 3% nonsense mutations, and 0.3% duplications (table 1).

Table 2 summarises all currently known PEX1 mutations as well as the available clinical phenotypes.

The most common PEX1 mutation, with an allele frequency of 0.43, is c.2528G→A, G843D in exon 15. The second most common mutation is the insertion c.2097_2098insT, I700YfsX42, with an allele frequency of 0.35. Thus these two common PEX1 mutations comprise about 80% of all abnormal alleles in CG1. The vast majority of PEX1 mutations are distributed over the two AAA cassettes with the functional protein domains, D1 and D2, and the highly conserved Walker motifs (fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 Distribution of PEX1 mutations over the functional protein domains. The protein has two AAA domains (AAA) with two functional domains, D1 and D2, and the highly conserved Walker A and B motifs (A1, B1, A2, B2). Asterixes mark the two most common PEX1 mutations, c.2528G→A, G843D and c.2097_2098insT, 1700YfsX42; ¹ = the first mutation identified in the crystallised NTD. NTD, N-terminal domain.

c.2528G→A, G843D is localised in the second functional domain (D2). c.2097_2098insT, I700YfsX42 is localised before the second functional domain, resulting in a truncated protein containing only one of the two nucleotide binding folds. In this study we identified seven mutations that have not been described so far—namely, c.911_912delCT, S304CfsX4; c.2083_2085delATG, M695del; c.3691_3694delCAGT, Q1231HfsX3; c.3287C→G, S1096X; c.274G→C, V92L; c.2387T→C, L796P; and c.3083G→A, R1013H. c.274G→C. V92L is the first mutation that has been detected in the crystallised NTD.

Mutations in the PEX1 gene are the most common cause of peroxisome biogenesis disorders.⁶ Alterations in this gene can lead to different clinical phenotypes comprising a continuum from severely affected patients with classical Zellweger syndrome through neonatal adrenoleucodystrophy, to infantile Refsum disease as the least severe. A summary of the data on genotype–phenotype correlations, including those already published and those derived from our own study (tables 2 and 3), confirms that there is a close link between the cellular impact of particular PEX1 mutations and the severity of the clinical course.

Class I mutations result in a peroxin 1 with residual protein levels and function, while class II mutations abolish PEX1 protein levels and function. Class I mutations are missense mutations, with c.2528G→A, G843D being by far the most frequent allele. c.2528G→A, G843D is a temperature sensitive mutation leading to reduced protein levels and function.^25,31 c.2528G→A, G843D on both alleles results in PEX1 protein levels of 3 to 20% compared to wild type. Other missense mutations like c.2392C→G, R798G that affects the first conserved functional protein domain D1 shows comparable PEX 1 protein levels and activity.³¹ Interaction of PEX1 with the second AAA-ATPase involved in peroxisome biogenesis, PEX6, in patients carrying c.2528G→A, G843D alleles is reduced by 70% when compared to wild type.³⁸ The most common class II mutation is c.2097_2098insT, I700YfsX42 resulting in a truncated PEX1 protein with a complete loss of the second functional domain, D2. In cells from CG1 patients bearing this mutation or other class II mutations including insertions, deletions, and splicing mutations, there is a total or nearly total loss of protein levels and function.^25,30 Patients with a class I mutation on both alleles present with a milder disease course, while those with class II mutation have a severe clinical phenotype. Patients who are compound heterozygotes for a class I and class II mutation (for example, c.2528G→A, G843D / c.2097_2098insT, I700YfsX42; c.2071+1G→T / c.1777G→A, G593R) show an intermediate clinical phenotype.^25,26

The incomplete knowledge of peroxisome biogenesis in general and of the pathophysiology in Zellweger spectrum patients and other PBDs hampers the attempts at devising effective treatments. The combined approach of functional studies in vitro and in mouse models with conditional alleles in vivo, as well as studying human cellular and clinical phenotypes, should further improve our understanding of the role of peroxins including PEX1 protein in peroxisome assembly. As the milder PBD phenotypes correlate with an unstable mutant protein with some residual function, treatment directing towards the identification of peroxisome associated factors that regulate protein stability of mutant PEX1 proteins might be a useful approach to reduce the severity of the disease.

Conclusions

PEX1 mutations are the major primary cause of Zellweger spectrum. These conditions include, in decreasing order of clinical severity, Zellweger syndrome, neonatal adrenoleucodystrophy, and infantile Refsum disease. In a given patient the degree of phenotypic severity correlates closely with the functional consequences of the PEX1 mutation on the encoded protein. In clinical practice, having a molecular confirmation of the clinical and biochemical diagnosis allows prediction of the clinical course of the disease in individual PBD cases. Furthermore, knowing the PEX1 mutation in an affected family is the only way to identify carriers. It ensures accurate genetic counselling and prenatal diagnosis on DNA based mutation detection techniques to identify the genotype of a fetus at risk.

DATABASES

PEX1 - OMIM: No 214100, 602136; GenBank: AC000064 (Homo sapiens, BAC clone), AF026086, AF030356, AB008112 and BC035575 (Homo sapiens, mRNA), AB052090 (Homo sapiens, mRNA for Pex1p-634del690), AB052091 (Homo sapiens, mRNA for Pex1pL664P), AB052092 (Homo sapiens, mRNA for Pex1pG843D), AB052093 (Homo sapiens, mRNA for Pex1pR633Ter), AB052094 (Homo sapiens, mRNA for Pex1pQ261Ter), CH236949 (Homo sapiens, genomic sequence).